EP0122048A2 - Datenverarbeitungszelle und Datenverarbeitungseinrichtungen mit solchen Zellen - Google Patents

Datenverarbeitungszelle und Datenverarbeitungseinrichtungen mit solchen Zellen Download PDF

Info

Publication number
EP0122048A2
EP0122048A2 EP84301600A EP84301600A EP0122048A2 EP 0122048 A2 EP0122048 A2 EP 0122048A2 EP 84301600 A EP84301600 A EP 84301600A EP 84301600 A EP84301600 A EP 84301600A EP 0122048 A2 EP0122048 A2 EP 0122048A2
Authority
EP
European Patent Office
Prior art keywords
data
signals
signal
input
output terminal
Prior art date
Legal status (The legal status is an assumption and is not a legal conclusion. Google has not performed a legal analysis and makes no representation as to the accuracy of the status listed.)
Granted
Application number
EP84301600A
Other languages
English (en)
French (fr)
Other versions
EP0122048A3 (en
EP0122048B1 (de
Inventor
Wlodzimierz Holsztynski
Current Assignee (The listed assignees may be inaccurate. Google has not performed a legal analysis and makes no representation or warranty as to the accuracy of the list.)
Martin Marietta Corp
Original Assignee
Martin Marietta Corp
Priority date (The priority date is an assumption and is not a legal conclusion. Google has not performed a legal analysis and makes no representation as to the accuracy of the date listed.)
Filing date
Publication date
Application filed by Martin Marietta Corp filed Critical Martin Marietta Corp
Publication of EP0122048A2 publication Critical patent/EP0122048A2/de
Publication of EP0122048A3 publication Critical patent/EP0122048A3/en
Application granted granted Critical
Publication of EP0122048B1 publication Critical patent/EP0122048B1/de
Expired legal-status Critical Current

Links

Images

Classifications

    • G—PHYSICS
    • G06—COMPUTING; CALCULATING OR COUNTING
    • G06F—ELECTRIC DIGITAL DATA PROCESSING
    • G06F15/00—Digital computers in general; Data processing equipment in general
    • G06F15/76—Architectures of general purpose stored program computers
    • G06F15/80—Architectures of general purpose stored program computers comprising an array of processing units with common control, e.g. single instruction multiple data processors
    • G06F15/8007—Architectures of general purpose stored program computers comprising an array of processing units with common control, e.g. single instruction multiple data processors single instruction multiple data [SIMD] multiprocessors
    • G06F15/8023—Two dimensional arrays, e.g. mesh, torus

Definitions

  • This invention relates to data processing cells and parallel data nrocessors incorporating such cells.
  • Parallel data processors typically employ a number of identical processing cells arranged in a matrix-like array so that data can be processed in parallel.
  • the usual parallel processor comprises n by m identical cells each connected to its immediately neighboring cells and each controlled by a central controller. This approach is particularly suited for processing data matrices of the types encountered in high resolution image processing.
  • a description of one such processing array and its many applications in data processing is set forth in US Patent No. 4 215 401 (Holsztynski et al).
  • each cell of the processor described in US 4 215 401 includes a random access memory (RAM), a single bit input accumulator, a single bit output accumulator and a NAND gate processing element.
  • the RAM, the input terminal of the input accumulator, the output terminal of the output accumulator and one input terminal of the NAND gate are connected to a data bus.
  • the output signal from the input accumulator is supplied to a second input terminal of the NAND gate which has its output terminal connected to the input terminal of the output accumulator.
  • the parallel processor described in US 4 215 401 can be programmed to rapidly perform a variety of elemental logical functions on large amounts of data. Combinations of these elemental functions by proper program sequences allow the processor to perform more complex logical and even arithmetic functions.
  • a parallel data processor having a plurality of cells for processing digital data signals, the processor being characterised by:
  • the invention also provides a data processing cell for processing single bit, digital signals, the cell being characterised by:
  • Preferred embodiments of the present invention described in detail hereinbelow provide: a parallel data processing method and system which performs both logical and arithmetic operations at very high speeds and yet is relatively simple, particularly at the processor cell level; a parallel data processing method and system which operates with relatively simple micro-instructions to perform arithmetic and logical functions with greater efficiency than known parallel processors; a bit processing method and cell which is of relatively simple construction, has a minimum of components and can rapidly perform arithmetic and logical functions; and a data processing method and processing cell which uses an arithmetic processor, in particular a full adder, to perform both arithmetic and logical operations.
  • the preferred processors which can be considered as geometric-arithmetic parallel data processors, comprise an arrangement of processing cells controlled by a central controller.
  • Each processing cell includes an arithmetic device connected to memories and signal selection circuits such that the arithmetic device can be controlled to perform both logical and arithmetic operations on data bits.
  • An array of identical cells is constructed by providing connections between each cell and its neighboring cells with edge cells in the array also communicating with suitable data input/output equipment.
  • a controller supplies control and address signals to the cells in accordance with a programmed sequence which results in the performance of desired operations on a data matrix introduced at the edge cells.
  • the arithmetic device in each processor cell is preferably a full adder.
  • a storage device such as a one bit register is connected to each input terminal of the full adder.
  • a storage device such as a one bit register is connected to each input terminal of the full adder. Since the adder has two data input terminals and a carry input terminal, three such storage devices are provided. One of the storage devices is designated the north/south (NS) input register, another is designated the east/west (EW) input register and the third is designated the carry (C) input register.
  • NS north/south
  • EW east/west
  • C carry
  • a signal selection circuit at the input terminal of each storage device selectively supplies input data to each storage device in response to control signals from the central controller.
  • the signals available for selection by the NS input register data selection circuit are the respective north and south output signals from the northern and southern neighboring cells, the output signal from the NS register itself, and a data bus signal.
  • the EW input register data selection circuit may choose from the respective east and west output signals from the eastern and western neighboring cells, the output signal from the EW register and the data bus signal.
  • the carry input register data selection circuit include the "carry" output signal from the full adder, the carry register output signal and a logic "one" and a logic "zero', preferably from the controller, or from some other suitable source.
  • a random access memory (RAM) of each cell selectively addressable by the controller, has its output terminal connected to a signal selection circuit which applies the input signals selected by the controller to the data bus.
  • the adder "sum” or plus output signal is also applied to this selection circuit for selective application to the bus in response to commands from the controller.
  • the RAM may be controlled by the controller to store or "write” data from the data bus and, if desired by addition of signal selection circuitry at the RAM data input terminal, to write data from other sources.
  • the RAM may be left unchanged when executing instructions not involving writing into the RAM.
  • data is selectively presented to the full adder to perform both logical and arithmetic functions. For example, by preloading a binary ONE in the carry register connected to the carry input terminal of the full adder and taking the output signal from the carry output terminal of the full adder, the logical OR of the inputs to the full adder data input terminals results. Similarly, by preloading a binary ZERO in the carry register, the carry output terminal of the full adder provides the logical AND of the data input signals.
  • transfer (routing) functions and arithmetic functions as well as other logical functions similarly can be commanded by relatively simple program routines.
  • a parallel processor implemented as an array of identical cells and embodying the present invention is functionally illustrated in the block diagram of Figure 1.
  • the edge cells of the array i.e., the cells in the first and last rows POO-PnO, Pm-Pnm and in the first and last columns POO-POm and Pn0-Pnm
  • Every cell receives command and address signals c 0 -c x and a 0 -a y , respectively, from a controller 12.
  • clock signals (not illustrated) may be provided from the controller 12 to each cell, and the data input/output unit may be controlled by the controller 12.
  • Each interior cell is connected to each of its neighboring cells so that four connections are provided to each cell.
  • the cells are each connected to their north, south, east, and west neighbors, the connections being designated N, S, E, and W, respectively, in Figure 1.
  • each cell can exchange data with any of its neighbors under the control of the controller 12.
  • input data to be processed is supplied to the data input/output unit 10.
  • This input data may be supplied, for example, from imaging devices which produce a large matrix of data "pixels", i.e., binary bits representing small parts of an image.
  • the input data supplied to the data input/output device 10, either directly or through the controller 12, are fed to the edge cells of the parallel processor which are controlled by the controller 12 to process this data in accordance with a sequence of micro-instructions.
  • the instructions may command processing which involves arithmetic operations, logical operations or merely data transfer.
  • the processing operation commanded by the controller 12 is simultaneously performed on all data by all cells.
  • the controller may command a series of simple transfers of data to neighboring cells in order to load an entire data matrix.
  • the controller 12 may command programmed sequence of operations to sum, subtract, multiply, complement, rotate, transpose or otherwise perform arithmetic and/or logical operations on individual bits in cell storage and/or from neighboring cells. The processing continues until all desired operations have been performed, and the results are transferred to the data input/output unit 10.
  • the data input/output unit 10 may be any suitable conventional electronic device for interfacing the parallel processor with other equipment.
  • this unit may include controllable input and output buffers for temporarily storing the input data before and after processing.
  • the controller 12 may be any suitable, conventional control unit capable of generating command signals and address signals in appropriate timed sequences in response to a sequence of program instructions. Suitable data input/output devices and controllers are illustrated and described in detail in U.S. Patent 4,215,401 and in the other prior art discussed herein.
  • the cell P includes an arithmetic processing element 20, preferably a full adder of any suitable conventional configuration.
  • the full adder has a first data input terminal D1, a second data input terminal D2 and a carry input terminal C.
  • the output terminals of the adder 20 include a sum output terminal and a carry output terminal which provide a PLUS output signal and a CARRY output signal, respectively.
  • a plurality of memories are connected to the full adder and are controllable by the command and address signals c o -c x , (where x is 7 in the illustrated embodiment) and a o -ay (where y is 6 in the illustrated embodiment) from the controller 12 of Figure 1.
  • the illustrated arrangement of memories connected as shown to the full adder permits the full adder to perform both logical and arithmetic functions.
  • single bit registers or memories 22, 24, and 26 designated the north/south or NS register, the east/west or EW register and the carry or C register, respectively, have their output terminals connected to the respective D1, D2 and C input terminals of the full adder 20. Storage of a particular data bit in the registers 22, 24 and 26 is determined by suitable conventional signal selection circuits (bit multiplexers MUX) 28, 30 and 32 under the control of the controller 12.
  • a north/south signal selection circuit or multiplexer 28 receives four data signals NS, N, BUS and S. In order to select one of these four data signals for application to the NS register 22, two control lines supply respective control signals c 0 and c 1 from the controller 12 to the selection circuit 28.
  • the circuit 28 operates in a conventional manner such that when the control or command signals c 0 , c l are both binary ZERO (00), the NS signal is applied to the NS register 22.
  • the command signals c 0 , c l are binary ONE and binary ZERO (10)
  • the N input signal is supplied to the NS register.
  • the combination of binary ZERO and binary ONE (01) applies the BUS signal to the NS register, and the combination of two binary ONE's (11) applies the S signal to the NS register.
  • One of the four data input signals EW, W, BUS and E is selectively supplied through the EW selection circuit (MUX) 30 to the EW register in response to the command signals c 2 , c 3 in the same manner as described above.
  • one of the data signals CY, ZERO, CARRY and ONE is selectively applied through the C selection circuit (MUX) 32 to the C register 26 in response to the c 4 and c 5 command signals.
  • the PLUS output signal from the sum output terminal of the full adder 20 is applied to one data input terminal of a two input bus selection circuit (MUX) 34.
  • the CARRY output signal from the carry output terminal of the full adder 20 is supplied to one input terminal of the carry or C selection circuit 32 as was previously mentioned.
  • the bus selection circuit 34 operates in response to the c 7 command signal to apply one of the two input signals to a data BUS.
  • the data BUS is connected to each of the data selection or multiplexer circuits 28, 30 and 38 to supply the BUS signal thereto.
  • An addressable multi-bit memory 36 preferably a 128 bit random access memory (RAM), is supplied with address signals a o -ay (wherein y is 6 for a 128 bit RAM) from the controller 12 of Figure 1.
  • the RAM output signal from the RAM 36 is supplied to the second input terminal of the bus selection circuit 34, while the input signal to the RAM 36 is either the BUS signal or the RAM output signal as selected by a RAM selection circuit (MUX) 38 under the control of the c 6 command signal.
  • MUX RAM selection circuit
  • the data input signals to each processor cell are as follows.
  • the N signal supplied to the NS selection circuit 28 is the signal from the NS register of the northern neighboring cell while the S signal supplied to circuit 28 is the signal from the NS register of the southern neighboring cell.
  • the W signal supplied to the EW selection circuit 30 is the signal from the EW register of the western neighboring cell while the E signal supplied to circuit 30 is the signal from the EW register of the eastern neighboring cell. It will thus be seen that each cell receives and'can store data bits from any neighboring cell (or from the data input/output unit in the case of edge cells).
  • the output signals from each processor cell are taken from the NS and EW registers.
  • the NS register output signal is designated as both the N and S output signals which are supplied to the NS selection circuits of southern and northern neighboring cells.
  • the EW register output signal is designated as both the E and W output signals which are supplied to the EW selection circuits of the respective western and eastern neighboring cells.
  • the NS output signal is also supplied to the NS selection circuit 28 of the immediate cell, and the EW output signal is supplied to the EW selection circuit 30 of the immediate cell.
  • Table I below systematically illustrates the manner in which data is transferred throughout each cell by commands from the controller 12 of Figure 1.
  • Table I describes the complete "architecture" of the cell of Figure 2 in that the commands which control all data transfers to a memory or to the bus are provided.
  • Table II below provides examples of full Commands which may be various combinations of the above basic commands used to simultaneously perform a number of data transfers. Not all possible commands are listed in Table II and it will be appreciated that some combinations will be illegal since it is not possible, for example, to simultaneously load and read the RAM. Thus, for example, the combination of 1, 0 for c 6 , c 7 may produce ambiguous results and should not be used.
  • the above basic actions commanded by the command words may be written in a micro-instruction notation for ease of understanding.
  • the complete instructions such.as those in Table II may be written as catenations of the basic instructions in Table I in the foregoing notation.
  • each program instruction is a catenation of some mnemonics, called “simple instructions”, subjected to some rules.
  • simple instructions subjected to some rules.
  • a program instruction is a catenation of the above "simple instructions” separated by blanks followed by a blank and double semicolon ;; preceded by an integer (the RAM address) if "RAM" appears in the catenation, and subjected to the following restrictions:
  • processor cell Since the processor cell has the above- described ability to transfer and store data, one skilled in the art will appreciate the ability of the processor to perform both arithmetic and logical functions in response to relatively simple sequences of micro instructions. To aid in understanding the operation of the processor cell in response to program instructions, it should first be noted that a full - adder operates in accordance with a truth table set forth below in Table III:
  • the program instructions of Table IV are each preceded by statements which explain the procedure to follow and the conditions which must be set up before proceeding with the routine. These statements, which are not part of the routines actually executed, are preceded by a percentage sign (%).
  • the part A of Table IV provides a generalized procedure and two more specific procedures (Al, A2 and A3, respectively) for loading a constant into a RAM plane.
  • routine Al the routine must be provided with two integers residing in the STOIC stack as the top two items of data. The first integer provided is a RAM address and the second is a constant "0" or "I". These integers will reside in the stack, respectively, next to the top (RAM address) and at the top (constant 0 or 1).
  • the actual program routine begins with the name of the routine after the apostrophe (') and ends after the last instruction at the single semi-colon (;).
  • the first of the two instructions loads PLUS into the NS and EW registers and loads the constant (on-the top of the STOIC stack, either 0 or 1) into the C register.
  • the PLUS output signal is then loaded into the RAM address location specified as the second item of data in the STOIC stack.
  • Figure 3 graphically illustrates a simple 6 by 6 data matrix in which each darkened square represents a binary ONE bit at a particular location in the matrix while each white square represents a binary ZERO bit at a particular location.
  • the parallel processor is a 6 by 6 array of identical cells and.that, ⁇ at the start of the processing cycle, each bit of the data matrix is loaded in the RAM of the appropriate cell in address position 5.
  • the 5th cell in the 2nd row (P52), the 4th and 5th cells in the 3rd row (P43, P53), the 3rd and 4th cells in the 4th row (P34, P44) and the 2nd cell in the 5th row (P25) of the processor array having binary ONE's in the RAM at address position 5. All other cells have a binary ZERO in this RAM address position.
  • FIG 4 graphically illustrates the result which would be obtained if the data illustrated in Figure 3 were subjected to a simple, exemplary conventional "erosion" process that is used extensively in image processing. It can be seen in Figure 4 that the result of the illustrated erosion process is a binary ONE in the 4th and 3rd cells of the respective 3rd and 4th rows (cells P43 and P34). Tb achieve this result, each data bit in each matrix location is compared with each bit to its right in the same row. If each of the two successive bits is a binary ONE, a binary ONE is stored in the RAM of the cell corresponding to the cell location from which the comparison is initiated (the "base" cell).
  • FIGS. 3 and 4 are graphic illustrations of the contents of RAM address 5 and 7 before and after erosion processing.
  • Instruction 1 loads both the EW and NS registers with the data bit in RAM address location 5 and sets the carry register to binary zero.
  • Instruction 2 loads the EW register with the data from the EW register of the eastern neighboring cell, (the E data signal), in effect shifting the RAM data in address location 5 to the left into the EW register of each western neighboring cell.
  • the data from RAM address 5 of each "base cell” is in the NS register of that cell and the data from the EW register of the eastern neighbor of each base cell is in the EW register of the base cell. Since the carry register has been set to ZERO, the CARRY output signal from the carry output terminal of the full adder is the AND logical function of the signals in the EW and NS registers (see subroutine C above).
  • the CARRY output signal is a binary ONE if the contents of both the NS and EW registers are binary ONE's, and is a binary ZERO for all other conditions.
  • the CARRY output signal-of each cell adder is the desired result, the AND function of each base cell and its immediate neighbor to the right (eastern neighbor).
  • the CARRY signal is stored in the carry register and both the NS and EW register are'set to the same value such as PLUS signal (instruction 3).
  • the PLUS signal then becomes the same as the signal in the carry register and can be stored in RAM by gating it through the bus selector and RAM selector circuits to the RAM input terminal and "writing" this signal into RAM memory location 7 (instruction 4).
  • Table VI Another data processing example using the processor cell of the Figure 2 embodiment is provided in Table VI below.
  • Table VI Another data processing example using the processor cell of the Figure 2 embodiment is provided in Table VI below.
  • Table VI Another data processing example using the processor cell of the Figure 2 embodiment is provided in Table VI below.
  • Table VI Another data processing example using the processor cell of the Figure 2 embodiment is provided in Table VI below.
  • Table VI Using the program of Table VI, a form of a convolution process is performed. Specifically, two neighboring points of data are, in essence, averaged by this program.
  • the program is written in STOIC.
  • Table VII provides another exemplary program for a processor using the processor cell of Figure 2 to further illustrate the capabilities of the cell.
  • the program of Table VII is a STOIC program which shows the ability of the cell of Figure 2 to perform subtractions.
  • Figure 5 illustrates one cell embodiment in which the full adder is expanded to include a "borrow" output signal BORROW and the NS and EW signals may be gated directly onto the data BUS.
  • control signals are provided such that a binary ZERO and ONE may be loaded directly into the NS and EW registers, eliminating the need for subroutines of instructions such as 7 above.
  • a CM register for direct south to north data shifts is provided.
  • a conventional one bit full adder/subtractor 40 having data input terminals Dl and D2 and a carry input terminal C, provides a PLUS output signal, a CARRY output signal and a BORROW output signal.
  • Single bit data accumulators or - registers 42, 44 and 46 designated the NS register, the EW register and the C register, respectively, have their output terminals connected to the respective Dl, D2 and C input terminals of the adder 40.
  • registers 42, 44 and 46 are selectively loaded, when clocked, from conventional signal controllable, signal selection circuits (multiplexers MUX) 48, 50 and 52, respectively, each of which can selectively apply any one of five input signals to its associated register in response to command signals from the controller 12 of Figure 1.
  • a 5-input signal selection circuit (MUX) 54 selectively applies one of five input signals to a data bus BUS, and a 3-input signal selection circuit (MUX) 56 selectively applies one of three input signals to the data input terminal of a single bit accumulator or register 58 designated the CM register.
  • a conventional 128 by 1 RAM 60 has the data BUS connected to its data input terminal DI.
  • the address inputs of the RAM receive the seven address signals a 0 -a 6 from the controller 12 of Figure 1, and the RAM output terminal DO is connected to one input terminal of the bus multiplexer (B MUX) 54.
  • a write enable input terminal of the RAM is controlled by the output signal from a conventional logic gate 62 which receives the c 9 , c 10 and clock (CLK) signals at its input terminals.
  • the NS signal selection circuit (NS MUX) 48 selects among the output signal from the NS register 42, the BUS signal, the N and S signals from the northern and southern neighboring cells, respectively and a control signal C15 which can be selectively set at ONE or ZERO by the controller.
  • the EW signal selection circuit (EW MUX) 50 selects among the EW register 44 output signal, the E and W signals from the respective eastern and western neighboring cells, the BUS signal and a c 15 command from the controller 12, all in response to the c 2 , c 3 and c 12 command signals from the controller.
  • the carry signal selection circuit (C MUX) 52 responds to the c 4 , c 5 and c 13 command signals from the controller 12 to select one of the CARRY, BORROW, BUS, c 15 and C signals, while the bus signal selection circuit (B MUX) 54 responds to the c 6 , c 7 and c 9 command signals to select from the NS register output signal, the EW register output signal, the PLUS signal, the CM register output signal and the RAM output signal for application to a data bus BUS.
  • the cell illustrated in Figure 5 operates in the same manner as that shown in Figure 2 except some of the functions may be accomplished with much simpler programs. For example, subtraction may be accomplished directly by the Figure 5 cell because of the provision of the borrow output signal from the full adder. With the full adder of Figure 2, the intermediate step of complementing one signal was necessary for subtraction. In addition, data shifts through the CM registers from south to north neighboring cells may be accomplished without interrupting ongoing processing. Thus, for example, an entire matrix may be loaded in the CM registers while an earlier loaded matrix is being processed.
  • the Figure 5 embodiment permits the programmer to directly load a ZERO or ONE in the NS and EW registers by control of the C15 command signal. This eliminates the need for a subroutine such as that set forth in instruction listing 7 above. Moreover, since the C register selection circuit 52 can be loaded directly from the BUS, subroutines such as 9 above can be simplified so as to place the NS signal on the BUS and then load the NS signal directly into the C register. Other programming simplifications rendered possible by the cell arrangement of Figure 5 will be appreciated by those skilled in the art to which the invention pertains.
  • the memories of a processor constructed as an array of cells embodying the present invention may be illustrated as memory "planes" to further facilitate an understanding of the operation of the processor.
  • the CM memories or registers of the Figure 5 embodiment are shown as a memory plane with the ability to directly shift data within the plane in the north direction but in no other direction.
  • the NS memories or registers of both cell embodiments are pictured as a memory plane with the ability to directly shift data in the north and south directions.
  • the EW memories or registers of both cell embodiments form a memory plane within which data can be directly shifted east and west.
  • the memory planes formed by the C register and the RAM do not have any direct data shifting capabilities.
  • the C register memory plane and the 128 memory planes of the RAM's are illustrated without direct shifting capability within the memory planes.
  • data from the C register memory plane and the RAM memory planes may be shifted from one cell to another, but not directly.
  • Intermediate steps such as first shifting this data into one of the direct shifting memory planes (NS, EW or, with the Figure 5 embodiment, CM) are required.
  • a parallel processor implemented with a cell arrangement embodying the -present invention is amenable to large scale integrated circuit (LSI) construction. With such construction there will be some array shape which will minimize the number of pins for a maximum number of cells.
  • LSI large scale integrated circuit
  • the rectangular array which minimizes the number of communication pins on an LSI chip per cell on the chip when implementing the Figure 5 embodiment for example, is one which has twice as many cells in the "m" or vertical (North/South) direction than in the horizontal "n" (East/West) direction.
  • the number of pins can be minimized for a maximum number of cells on an LSI chip by providing a rectangular matrix or array of cells with equal numbers of cells in each direction.
  • the 8 by 8 array would have a minimum number of pins for a maximum array size.
  • VLSI integrated circuit
  • the part of the cell comprising the processing element of the previous - embodiments is illustrated using a conventional VLSI "lattice" diagram.
  • the signals shown on the left side of the Figure are the control signals and memory signals previously discussed, although it should be noted that both a signal and its complement are provided in most instances (e.g., the EW/and EW signals).
  • the vertical lines in Figure 7 represent logical AND functions of the signals supplied from the left on the horizontal lines. More specifically, if one follows a vertical line and finds a dot at the intersection of two horizontal lines, the signal on the vertical line is the AND function of those two signals coming in on the horizontal lines.
  • the signal on the top horizontal line is a system clock which is applied to each vertical line.
  • the vertical lines also are interconnected in groups at the bottom of the Figure. Each such group connection represents the logical OR of all the signals appearing on the vertical lines. The output signals from these groups of vertical lines are shown at the bottom of Figure 7.
  • each group of "OR'ed" vertical lines has four lines which provide the AND of C/,NS,EW and C,NS,EW/ and C, NS/, EW and C/,NS/,EW/ which together perform the adder/subtractor truth table functions.
  • the bus function is performed within the logic array.
  • FIG. 1 While one configuration of a processor using cells embodying the present invention is shown in Figure 1 as a rectangular array of m by n cells where m and n are both greater than one, other configurations may be desirable.
  • Figure 8 One possible configuration for such a linear processor is illustrated in Figure 8.
  • a plurality of identical processor cells Pl-Pn are connected so that the east terminal E of each cell is connected to the west terminal W of each neighboring cell and the south terminal S of each cell is connected to the north terminal N of its neighboring cell.
  • the first and last cells Pl and Pn have their respective north and west terminals N, W and south and east terminals S, E connected to the data input/output unit 10'.
  • Each cell receives the command signals c o -c x and the address signals a o -ay from the controller 12'.
  • the data input/output unit is like that previously described in connection with Figure 1 except it has only one input/output line for each of the N, S, E and W signals since there is only one row of cells.
  • the controller similarly is like that described in connection with Figure 1.
  • Linear data is loaded into the processor by shifting it out of the input/output unit 12' into the first cell and then to subsequent cells until a full line of data is stored in the processor.
  • an entire line of data may be shifted into the processor in parallel with a slightly modified, parallel connection between the processor and the input/output unit.
  • the processor may be controlled in any desired manner as was previously described to perform arithmetic and/or logical functions on the data. The processed data is then available for shifting to the input/output unit for further use.

Landscapes

  • Engineering & Computer Science (AREA)
  • Computer Hardware Design (AREA)
  • Theoretical Computer Science (AREA)
  • Computing Systems (AREA)
  • Physics & Mathematics (AREA)
  • General Engineering & Computer Science (AREA)
  • General Physics & Mathematics (AREA)
  • Complex Calculations (AREA)
  • Multi Processors (AREA)
  • Image Processing (AREA)
EP84301600A 1983-03-10 1984-03-09 Datenverarbeitungszelle und Datenverarbeitungseinrichtungen mit solchen Zellen Expired EP0122048B1 (de)

Applications Claiming Priority (2)

Application Number Priority Date Filing Date Title
US06/474,209 US4739474A (en) 1983-03-10 1983-03-10 Geometric-arithmetic parallel processor
US474209 1983-03-10

Publications (3)

Publication Number Publication Date
EP0122048A2 true EP0122048A2 (de) 1984-10-17
EP0122048A3 EP0122048A3 (en) 1987-01-07
EP0122048B1 EP0122048B1 (de) 1990-11-07

Family

ID=23882614

Family Applications (1)

Application Number Title Priority Date Filing Date
EP84301600A Expired EP0122048B1 (de) 1983-03-10 1984-03-09 Datenverarbeitungszelle und Datenverarbeitungseinrichtungen mit solchen Zellen

Country Status (5)

Country Link
US (1) US4739474A (de)
EP (1) EP0122048B1 (de)
JP (1) JP2671120B2 (de)
CA (1) CA1201208A (de)
DE (1) DE3483541D1 (de)

Cited By (7)

* Cited by examiner, † Cited by third party
Publication number Priority date Publication date Assignee Title
US4698243A (en) * 1986-06-20 1987-10-06 The Dow Chemical Company Method for sizing and hydrolyzing polytetrafluoroethylene fabrics, fibers, yarns, or threads
US4778723A (en) * 1986-06-20 1988-10-18 The Dow Chemical Company Method for sizing polytetrafluoroethylene fibers, yarn, or threads
US4783738A (en) * 1986-03-13 1988-11-08 International Business Machines Corporation Adaptive instruction processing by array processor having processor identification and data dependent status registers in each processing element
GB2204431A (en) * 1987-04-25 1988-11-09 Ferranti Plc Digital signal processing and transducer array beamforming
US4784900A (en) * 1985-05-31 1988-11-15 University Of Bath Method for sizing polytretrafluoroethylene fabrics
WO1993011503A1 (en) * 1991-12-06 1993-06-10 Norman Richard S Massively-parallel direct output processor array
US5757641A (en) * 1995-07-03 1998-05-26 General Electric Company Triplex control system with sensor failure compensation

Families Citing this family (126)

* Cited by examiner, † Cited by third party
Publication number Priority date Publication date Assignee Title
DE3506749A1 (de) * 1984-02-27 1985-09-26 Nippon Telegraph & Telephone Public Corp., Tokio/Tokyo Matrixprozessor und steuerverfahren hierfuer
DE3533800C1 (de) * 1985-09-21 1987-02-05 Hans-Werner Lang Verfahren zum Betreiben eines hochintegrierten Wellenfront-Feldrechners sowie entsprechender Wellenfront-Feldrechner
US4860249A (en) * 1985-10-04 1989-08-22 Saxpy Computer Corporation Multizone array processor implementing two sided zone buffers with each side being dynamically configured as a working or I/O side
US4775952A (en) * 1986-05-29 1988-10-04 General Electric Company Parallel processing system apparatus
US4985832A (en) * 1986-09-18 1991-01-15 Digital Equipment Corporation SIMD array processing system with routing networks having plurality of switching stages to transfer messages among processors
DE3804938C2 (de) 1987-02-18 1994-07-28 Canon Kk Bildverarbeitungseinrichtung
US5293481A (en) * 1987-02-18 1994-03-08 Canon Kabushiki Kaisha Data parallel processing apparatus
US5058001A (en) * 1987-03-05 1991-10-15 International Business Machines Corporation Two-dimensional array of processing elements for emulating a multi-dimensional network
US4891751A (en) * 1987-03-27 1990-01-02 Floating Point Systems, Inc. Massively parallel vector processing computer
US5129092A (en) * 1987-06-01 1992-07-07 Applied Intelligent Systems,Inc. Linear chain of parallel processors and method of using same
US4908751A (en) * 1987-10-15 1990-03-13 Smith Harry F Parallel data processor
US4951221A (en) * 1988-04-18 1990-08-21 General Electric Company Cell stack for variable digit width serial architecture
US5257395A (en) * 1988-05-13 1993-10-26 International Business Machines Corporation Methods and circuit for implementing and arbitrary graph on a polymorphic mesh
WO1990004235A1 (en) * 1988-10-07 1990-04-19 Martin Marietta Corporation Parallel data processor
US5179714A (en) * 1988-10-07 1993-01-12 Martin Marietta Corporation Parallel bit serial data processor
US5253308A (en) * 1989-06-21 1993-10-12 Amber Engineering, Inc. Massively parallel digital image data processor using pixel-mapped input/output and relative indexed addressing
CA2021192A1 (en) * 1989-07-28 1991-01-29 Malcolm A. Mumme Simplified synchronous mesh processor
US5173947A (en) * 1989-08-01 1992-12-22 Martin Marietta Corporation Conformal image processing apparatus and method
US5452425A (en) * 1989-10-13 1995-09-19 Texas Instruments Incorporated Sequential constant generator system for indicating the last data word by using the end of loop bit having opposite digital state than other data words
US5327541A (en) * 1989-10-13 1994-07-05 Texas Instruments Inc. Global rotation of data in synchronous vector processor
US5765010A (en) * 1989-10-13 1998-06-09 Texas Instruments Incorporated Timing and control circuit and method for a synchronous vector processor
US5163120A (en) * 1989-10-13 1992-11-10 Texas Instruments Incorporated Second nearest-neighbor communication network for synchronous vector processor, systems and methods
US5408673A (en) * 1989-10-13 1995-04-18 Texas Instruments Incorporated Circuit for continuous processing of video signals in a synchronous vector processor and method of operating same
US5293637A (en) * 1989-10-13 1994-03-08 Texas Instruments Distribution of global variables in synchronous vector processor
US5105387A (en) * 1989-10-13 1992-04-14 Texas Instruments Incorporated Three transistor dual port dynamic random access memory gain cell
US5598545A (en) * 1989-10-13 1997-01-28 Texas Instruments Incorporated Circuitry and method for performing two operating instructions during a single clock in a processing device
US5539891A (en) * 1989-10-13 1996-07-23 Texas Instruments Incorporated Data transfer control circuit with a sequencer circuit and control subcircuits and data control method for successively entering data into a memory
US5210836A (en) * 1989-10-13 1993-05-11 Texas Instruments Incorporated Instruction generator architecture for a video signal processor controller
US5321510A (en) * 1989-11-13 1994-06-14 Texas Instruments Incorporated Serial video processor
GB8925723D0 (en) * 1989-11-14 1990-01-04 Amt Holdings Processor array system
US5091786A (en) * 1990-03-01 1992-02-25 Texas Instruments Incorporated Multi-screen feature for improved definition television digital processing units, systems, and methods
US5091783A (en) * 1990-03-01 1992-02-25 Texas Instruments Incorporated Still more feature for improved definition television digital processing units, systems, and methods
US5093722A (en) * 1990-03-01 1992-03-03 Texas Instruments Incorporated Definition television digital processing units, systems and methods
US5133073A (en) * 1990-05-29 1992-07-21 Wavetracer, Inc. Processor array of N-dimensions which is physically reconfigurable into N-1
US5193202A (en) * 1990-05-29 1993-03-09 Wavetracer, Inc. Processor array with relocated operand physical address generator capable of data transfer to distant physical processor for each virtual processor while simulating dimensionally larger array processor
US5157785A (en) * 1990-05-29 1992-10-20 Wavetracer, Inc. Process cell for an n-dimensional processor array having a single input element with 2n data inputs, memory, and full function arithmetic logic unit
US5420754A (en) * 1990-09-28 1995-05-30 At&T Corp. Stacked board assembly for computing machines, including routing boards
US5963745A (en) * 1990-11-13 1999-10-05 International Business Machines Corporation APAP I/O programmable router
US5752067A (en) * 1990-11-13 1998-05-12 International Business Machines Corporation Fully scalable parallel processing system having asynchronous SIMD processing
US5966528A (en) * 1990-11-13 1999-10-12 International Business Machines Corporation SIMD/MIMD array processor with vector processing
US5617577A (en) * 1990-11-13 1997-04-01 International Business Machines Corporation Advanced parallel array processor I/O connection
US5765015A (en) * 1990-11-13 1998-06-09 International Business Machines Corporation Slide network for an array processor
US5794059A (en) * 1990-11-13 1998-08-11 International Business Machines Corporation N-dimensional modified hypercube
US5630162A (en) * 1990-11-13 1997-05-13 International Business Machines Corporation Array processor dotted communication network based on H-DOTs
US5963746A (en) * 1990-11-13 1999-10-05 International Business Machines Corporation Fully distributed processing memory element
US5625836A (en) * 1990-11-13 1997-04-29 International Business Machines Corporation SIMD/MIMD processing memory element (PME)
DE69131272T2 (de) * 1990-11-13 1999-12-09 Ibm Paralleles Assoziativprozessor-System
US5828894A (en) * 1990-11-13 1998-10-27 International Business Machines Corporation Array processor having grouping of SIMD pickets
US5809292A (en) * 1990-11-13 1998-09-15 International Business Machines Corporation Floating point for simid array machine
US5765011A (en) * 1990-11-13 1998-06-09 International Business Machines Corporation Parallel processing system having a synchronous SIMD processing with processing elements emulating SIMD operation using individual instruction streams
US5590345A (en) * 1990-11-13 1996-12-31 International Business Machines Corporation Advanced parallel array processor(APAP)
US5734921A (en) * 1990-11-13 1998-03-31 International Business Machines Corporation Advanced parallel array processor computer package
US5815723A (en) * 1990-11-13 1998-09-29 International Business Machines Corporation Picket autonomy on a SIMD machine
US5588152A (en) * 1990-11-13 1996-12-24 International Business Machines Corporation Advanced parallel processor including advanced support hardware
US5765012A (en) * 1990-11-13 1998-06-09 International Business Machines Corporation Controller for a SIMD/MIMD array having an instruction sequencer utilizing a canned routine library
US5325500A (en) * 1990-12-14 1994-06-28 Xerox Corporation Parallel processing units on a substrate, each including a column of memory
US5594918A (en) * 1991-05-13 1997-01-14 International Business Machines Corporation Parallel computer system providing multi-ported intelligent memory
DE69229146T2 (de) * 1991-06-25 1999-10-28 Canon Kk Verfahren und Vorrichtung zur Detektion eines Bewegungsvektors sowie Kodierungsverfahren und Vorrichtung zur Anwendung eines solchen Verfahrens und Vorrichtung
US5689719A (en) * 1991-06-28 1997-11-18 Sanyo Electric O., Ltd. Parallel computer system including processing elements
US5506992A (en) * 1992-01-30 1996-04-09 Saxenmeyer; George Distributed processing system with asynchronous communication between processing modules
JP2642039B2 (ja) * 1992-05-22 1997-08-20 インターナショナル・ビジネス・マシーンズ・コーポレイション アレイ・プロセッサ
US6298162B1 (en) 1992-12-23 2001-10-02 Lockheed Martin Corporation Image compression/expansion using parallel decomposition/recomposition
US6073185A (en) * 1993-08-27 2000-06-06 Teranex, Inc. Parallel data processor
US5535291A (en) * 1994-02-18 1996-07-09 Martin Marietta Corporation Superresolution image enhancement for a SIMD array processor
US5557734A (en) * 1994-06-17 1996-09-17 Applied Intelligent Systems, Inc. Cache burst architecture for parallel processing, such as for image processing
DE69424304T2 (de) * 1994-09-13 2000-11-30 Teranex Orlando Paralleler datenprozessor
KR100186918B1 (ko) * 1994-10-21 1999-05-01 모리시다 요이치 신호처리장치
US5943242A (en) * 1995-11-17 1999-08-24 Pact Gmbh Dynamically reconfigurable data processing system
US7266725B2 (en) * 2001-09-03 2007-09-04 Pact Xpp Technologies Ag Method for debugging reconfigurable architectures
US6388672B1 (en) * 1996-02-02 2002-05-14 Kabushiki Kaisha Toshiba Graphic translate engine, floating point arithmetic unit and floating point multiply-add calculation unit
DE19651075A1 (de) * 1996-12-09 1998-06-10 Pact Inf Tech Gmbh Einheit zur Verarbeitung von numerischen und logischen Operationen, zum Einsatz in Prozessoren (CPU's), Mehrrechnersystemen, Datenflußprozessoren (DFP's), digitalen Signal Prozessoren (DSP's) oder dergleichen
US6338106B1 (en) 1996-12-20 2002-01-08 Pact Gmbh I/O and memory bus system for DFPS and units with two or multi-dimensional programmable cell architectures
DE19654593A1 (de) * 1996-12-20 1998-07-02 Pact Inf Tech Gmbh Umkonfigurierungs-Verfahren für programmierbare Bausteine zur Laufzeit
DE19654595A1 (de) 1996-12-20 1998-07-02 Pact Inf Tech Gmbh I0- und Speicherbussystem für DFPs sowie Bausteinen mit zwei- oder mehrdimensionaler programmierbaren Zellstrukturen
DE59710317D1 (de) * 1996-12-27 2003-07-24 Pact Inf Tech Gmbh VERFAHREN ZUM SELBSTÄNDIGEN DYNAMISCHEN UMLADEN VON DATENFLUSSPROZESSOREN (DFPs) SOWIE BAUSTEINEN MIT ZWEI- ODER MEHRDIMENSIONALEN PROGRAMMIERBAREN ZELLSTRUKTUREN (FPGAs, DPGAs, o.dgl.)
DE19654846A1 (de) * 1996-12-27 1998-07-09 Pact Inf Tech Gmbh Verfahren zum selbständigen dynamischen Umladen von Datenflußprozessoren (DFPs) sowie Bausteinen mit zwei- oder mehrdimensionalen programmierbaren Zellstrukturen (FPGAs, DPGAs, o. dgl.)
DE19704044A1 (de) * 1997-02-04 1998-08-13 Pact Inf Tech Gmbh Verfahren zur automatischen Adressgenerierung von Bausteinen innerhalb Clustern aus einer Vielzahl dieser Bausteine
US6542998B1 (en) 1997-02-08 2003-04-01 Pact Gmbh Method of self-synchronization of configurable elements of a programmable module
DE19704728A1 (de) * 1997-02-08 1998-08-13 Pact Inf Tech Gmbh Verfahren zur Selbstsynchronisation von konfigurierbaren Elementen eines programmierbaren Bausteines
DE19704742A1 (de) 1997-02-11 1998-09-24 Pact Inf Tech Gmbh Internes Bussystem für DFPs, sowie Bausteinen mit zwei- oder mehrdimensionalen programmierbaren Zellstrukturen, zur Bewältigung großer Datenmengen mit hohem Vernetzungsaufwand
US8686549B2 (en) 2001-09-03 2014-04-01 Martin Vorbach Reconfigurable elements
DE19861088A1 (de) * 1997-12-22 2000-02-10 Pact Inf Tech Gmbh Verfahren zur Reparatur von integrierten Schaltkreisen
DE19807872A1 (de) 1998-02-25 1999-08-26 Pact Inf Tech Gmbh Verfahren zur Verwaltung von Konfigurationsdaten in Datenflußprozessoren sowie Bausteinen mit zwei- oder mehrdimensionalen programmierbaren Zellstruktur (FPGAs, DPGAs, o. dgl.
US6173388B1 (en) 1998-04-09 2001-01-09 Teranex Inc. Directly accessing local memories of array processors for improved real-time corner turning processing
US6212628B1 (en) 1998-04-09 2001-04-03 Teranex, Inc. Mesh connected computer
US6185667B1 (en) 1998-04-09 2001-02-06 Teranex, Inc. Input/output support for processing in a mesh connected computer
US6067609A (en) * 1998-04-09 2000-05-23 Teranex, Inc. Pattern generation and shift plane operations for a mesh connected computer
EP1228440B1 (de) 1999-06-10 2017-04-05 PACT XPP Technologies AG Sequenz-partitionierung auf zellstrukturen
DE50115584D1 (de) 2000-06-13 2010-09-16 Krass Maren Pipeline ct-protokolle und -kommunikation
JP2004517386A (ja) 2000-10-06 2004-06-10 ペーアーツェーテー イクスペーペー テクノロジーズ アクチエンゲゼルシャフト 方法および装置
US20040015899A1 (en) * 2000-10-06 2004-01-22 Frank May Method for processing data
US8058899B2 (en) 2000-10-06 2011-11-15 Martin Vorbach Logic cell array and bus system
US6990555B2 (en) * 2001-01-09 2006-01-24 Pact Xpp Technologies Ag Method of hierarchical caching of configuration data having dataflow processors and modules having two- or multidimensional programmable cell structure (FPGAs, DPGAs, etc.)
US20020133688A1 (en) * 2001-01-29 2002-09-19 Ming-Hau Lee SIMD/MIMD processing on a reconfigurable array
US9037807B2 (en) 2001-03-05 2015-05-19 Pact Xpp Technologies Ag Processor arrangement on a chip including data processing, memory, and interface elements
US7581076B2 (en) 2001-03-05 2009-08-25 Pact Xpp Technologies Ag Methods and devices for treating and/or processing data
US7444531B2 (en) 2001-03-05 2008-10-28 Pact Xpp Technologies Ag Methods and devices for treating and processing data
US7210129B2 (en) 2001-08-16 2007-04-24 Pact Xpp Technologies Ag Method for translating programs for reconfigurable architectures
US7844796B2 (en) * 2001-03-05 2010-11-30 Martin Vorbach Data processing device and method
EP2224330B1 (de) 2001-06-20 2012-05-09 Krass, Maren Verfahren und gerät zum partitionieren von grossen rechnerprogrammen
US7996827B2 (en) 2001-08-16 2011-08-09 Martin Vorbach Method for the translation of programs for reconfigurable architectures
US7434191B2 (en) 2001-09-03 2008-10-07 Pact Xpp Technologies Ag Router
US8686475B2 (en) 2001-09-19 2014-04-01 Pact Xpp Technologies Ag Reconfigurable elements
US7577822B2 (en) 2001-12-14 2009-08-18 Pact Xpp Technologies Ag Parallel task operation in processor and reconfigurable coprocessor configured based on information in link list including termination information for synchronization
EP1483682A2 (de) 2002-01-19 2004-12-08 PACT XPP Technologies AG Reconfigurierbarer prozessor
EP2043000B1 (de) 2002-02-18 2011-12-21 Richter, Thomas Bussysteme und Rekonfigurationsverfahren
US8914590B2 (en) 2002-08-07 2014-12-16 Pact Xpp Technologies Ag Data processing method and device
WO2004021176A2 (de) 2002-08-07 2004-03-11 Pact Xpp Technologies Ag Verfahren und vorrichtung zur datenverarbeitung
US20070083730A1 (en) * 2003-06-17 2007-04-12 Martin Vorbach Data processing device and method
US7657861B2 (en) 2002-08-07 2010-02-02 Pact Xpp Technologies Ag Method and device for processing data
AU2003289844A1 (en) 2002-09-06 2004-05-13 Pact Xpp Technologies Ag Reconfigurable sequencer structure
US7146486B1 (en) * 2003-01-29 2006-12-05 S3 Graphics Co., Ltd. SIMD processor with scalar arithmetic logic units
US20040252547A1 (en) * 2003-06-06 2004-12-16 Chengpu Wang Concurrent Processing Memory
JP4700611B2 (ja) * 2003-08-28 2011-06-15 ペーアーツェーテー イクスペーペー テクノロジーズ アクチエンゲゼルシャフト データ処理装置およびデータ処理方法
US7937557B2 (en) * 2004-03-16 2011-05-03 Vns Portfolio Llc System and method for intercommunication between computers in an array
US7904695B2 (en) 2006-02-16 2011-03-08 Vns Portfolio Llc Asynchronous power saving computer
US7281942B2 (en) * 2005-11-18 2007-10-16 Ideal Industries, Inc. Releasable wire connector
US8250503B2 (en) 2006-01-18 2012-08-21 Martin Vorbach Hardware definition method including determining whether to implement a function as hardware or software
US7904615B2 (en) * 2006-02-16 2011-03-08 Vns Portfolio Llc Asynchronous computer communication
US7966481B2 (en) * 2006-02-16 2011-06-21 Vns Portfolio Llc Computer system and method for executing port communications without interrupting the receiving computer
GB2436377B (en) * 2006-03-23 2011-02-23 Cambridge Display Tech Ltd Data processing hardware
JP2007272895A (ja) * 2006-03-31 2007-10-18 Technology Properties Ltd コンピュータプロセッサアレイの操作方法および装置
US20080270751A1 (en) * 2007-04-27 2008-10-30 Technology Properties Limited System and method for processing data in a pipeline of computers
US20080282062A1 (en) * 2007-05-07 2008-11-13 Montvelishsky Michael B Method and apparatus for loading data and instructions into a computer
WO2010011813A1 (en) * 2008-07-23 2010-01-28 Alkermes, Inc. Complex of trospium and pharmaceutical compositions thereof
US8898432B2 (en) 2011-10-25 2014-11-25 Geo Semiconductor, Inc. Folded SIMD array organized in groups (PEGs) of respective array segments, control signal distribution logic, and local memory

Citations (2)

* Cited by examiner, † Cited by third party
Publication number Priority date Publication date Assignee Title
GB1536933A (en) 1977-03-16 1978-12-29 Int Computers Ltd Array processors
US4215401A (en) 1978-09-28 1980-07-29 Environmental Research Institute Of Michigan Cellular digital array processor

Family Cites Families (5)

* Cited by examiner, † Cited by third party
Publication number Priority date Publication date Assignee Title
US3701976A (en) * 1970-07-13 1972-10-31 Bell Telephone Labor Inc Floating point arithmetic unit for a parallel processing computer
US3815095A (en) * 1972-08-29 1974-06-04 Texas Instruments Inc General-purpose array processor
US4065808A (en) * 1975-01-25 1977-12-27 U.S. Philips Corporation Network computer system
US4314349A (en) * 1979-12-31 1982-02-02 Goodyear Aerospace Corporation Processing element for parallel array processors
US4524455A (en) * 1981-06-01 1985-06-18 Environmental Research Inst. Of Michigan Pipeline processor

Patent Citations (2)

* Cited by examiner, † Cited by third party
Publication number Priority date Publication date Assignee Title
GB1536933A (en) 1977-03-16 1978-12-29 Int Computers Ltd Array processors
US4215401A (en) 1978-09-28 1980-07-29 Environmental Research Institute Of Michigan Cellular digital array processor

Non-Patent Citations (1)

* Cited by examiner, † Cited by third party
Title
JOHN H. SMIT: "Very High Speed Computing Symposium", 9 September 1980, GEORGIA INSTITUTE OF TECHNOLOGY, article "Architecture Description For The Massively Parallel Processor (MPP) And The Airborne Associative Processor (ASPRO)"

Cited By (7)

* Cited by examiner, † Cited by third party
Publication number Priority date Publication date Assignee Title
US4784900A (en) * 1985-05-31 1988-11-15 University Of Bath Method for sizing polytretrafluoroethylene fabrics
US4783738A (en) * 1986-03-13 1988-11-08 International Business Machines Corporation Adaptive instruction processing by array processor having processor identification and data dependent status registers in each processing element
US4698243A (en) * 1986-06-20 1987-10-06 The Dow Chemical Company Method for sizing and hydrolyzing polytetrafluoroethylene fabrics, fibers, yarns, or threads
US4778723A (en) * 1986-06-20 1988-10-18 The Dow Chemical Company Method for sizing polytetrafluoroethylene fibers, yarn, or threads
GB2204431A (en) * 1987-04-25 1988-11-09 Ferranti Plc Digital signal processing and transducer array beamforming
WO1993011503A1 (en) * 1991-12-06 1993-06-10 Norman Richard S Massively-parallel direct output processor array
US5757641A (en) * 1995-07-03 1998-05-26 General Electric Company Triplex control system with sensor failure compensation

Also Published As

Publication number Publication date
CA1201208A (en) 1986-02-25
EP0122048A3 (en) 1987-01-07
EP0122048B1 (de) 1990-11-07
JPS59172065A (ja) 1984-09-28
JP2671120B2 (ja) 1997-10-29
US4739474A (en) 1988-04-19
DE3483541D1 (de) 1990-12-13

Similar Documents

Publication Publication Date Title
EP0122048B1 (de) Datenverarbeitungszelle und Datenverarbeitungseinrichtungen mit solchen Zellen
EP0390907B1 (de) Paralleler datenprozessor
US4215401A (en) Cellular digital array processor
US3287703A (en) Computer
US5129092A (en) Linear chain of parallel processors and method of using same
KR100373222B1 (ko) 반도체집적회로
US3979728A (en) Array processors
US4380046A (en) Massively parallel processor computer
US6067609A (en) Pattern generation and shift plane operations for a mesh connected computer
US4622632A (en) Data processing system having a pyramidal array of processors
JP2562845B2 (ja) メッシュプロセッサアレー
US5247613A (en) Massively parallel processor including transpose arrangement for serially transmitting bits of data words stored in parallel
US5179714A (en) Parallel bit serial data processor
US5321510A (en) Serial video processor
US6185667B1 (en) Input/output support for processing in a mesh connected computer
EP0188059B1 (de) Halbleiterspeichereinrichtung mit Lese-Änderung-Schreib-Konfiguration
JPS62264357A (ja) 並列プロセツサ・アレイにおけるプロセツサ付加のシミユレ−シヨン方法
US4434502A (en) Memory system handling a plurality of bits as a unit to be processed
US20050024983A1 (en) Providing a register file memory with local addressing in a SIMD parallel processor
US4939642A (en) Virtual bit map processor
EP0557997A2 (de) Informationsverarbeitungsanlage und -system
US4939575A (en) Fault-tolerant serial video processor device
JPH09505921A (ja) 並列データ処理プロセッサ
US6175852B1 (en) High-speed binary adder
EP0317218B1 (de) Serieller Videoprozessor und Verfahren

Legal Events

Date Code Title Description
PUAI Public reference made under article 153(3) epc to a published international application that has entered the european phase

Free format text: ORIGINAL CODE: 0009012

AK Designated contracting states

Designated state(s): CH DE FR GB IT LI NL SE

PUAL Search report despatched

Free format text: ORIGINAL CODE: 0009013

AK Designated contracting states

Kind code of ref document: A3

Designated state(s): CH DE FR GB IT LI NL SE

17P Request for examination filed

Effective date: 19870528

17Q First examination report despatched

Effective date: 19881114

ITF It: translation for a ep patent filed

Owner name: ST. ASSOC. MARIETTI & PIPPARELLI

GRAA (expected) grant

Free format text: ORIGINAL CODE: 0009210

AK Designated contracting states

Kind code of ref document: B1

Designated state(s): CH DE FR GB IT LI NL SE

ET Fr: translation filed
REF Corresponds to:

Ref document number: 3483541

Country of ref document: DE

Date of ref document: 19901213

PLBE No opposition filed within time limit

Free format text: ORIGINAL CODE: 0009261

STAA Information on the status of an ep patent application or granted ep patent

Free format text: STATUS: NO OPPOSITION FILED WITHIN TIME LIMIT

26N No opposition filed
ITTA It: last paid annual fee
EAL Se: european patent in force in sweden

Ref document number: 84301600.7

PGFP Annual fee paid to national office [announced via postgrant information from national office to epo]

Ref country code: FR

Payment date: 20000202

Year of fee payment: 17

PGFP Annual fee paid to national office [announced via postgrant information from national office to epo]

Ref country code: CH

Payment date: 20000221

Year of fee payment: 17

PGFP Annual fee paid to national office [announced via postgrant information from national office to epo]

Ref country code: SE

Payment date: 20000222

Year of fee payment: 17

Ref country code: GB

Payment date: 20000222

Year of fee payment: 17

Ref country code: DE

Payment date: 20000222

Year of fee payment: 17

PGFP Annual fee paid to national office [announced via postgrant information from national office to epo]

Ref country code: NL

Payment date: 20000223

Year of fee payment: 17

PG25 Lapsed in a contracting state [announced via postgrant information from national office to epo]

Ref country code: GB

Free format text: LAPSE BECAUSE OF NON-PAYMENT OF DUE FEES

Effective date: 20010309

PG25 Lapsed in a contracting state [announced via postgrant information from national office to epo]

Ref country code: SE

Free format text: LAPSE BECAUSE OF NON-PAYMENT OF DUE FEES

Effective date: 20010310

PG25 Lapsed in a contracting state [announced via postgrant information from national office to epo]

Ref country code: LI

Free format text: LAPSE BECAUSE OF NON-PAYMENT OF DUE FEES

Effective date: 20010331

Ref country code: CH

Free format text: LAPSE BECAUSE OF NON-PAYMENT OF DUE FEES

Effective date: 20010331

PG25 Lapsed in a contracting state [announced via postgrant information from national office to epo]

Ref country code: NL

Free format text: LAPSE BECAUSE OF NON-PAYMENT OF DUE FEES

Effective date: 20011001

GBPC Gb: european patent ceased through non-payment of renewal fee

Effective date: 20010309

EUG Se: european patent has lapsed

Ref document number: 84301600.7

REG Reference to a national code

Ref country code: CH

Ref legal event code: PL

PG25 Lapsed in a contracting state [announced via postgrant information from national office to epo]

Ref country code: FR

Free format text: LAPSE BECAUSE OF NON-PAYMENT OF DUE FEES

Effective date: 20011130

NLV4 Nl: lapsed or anulled due to non-payment of the annual fee

Effective date: 20011001

REG Reference to a national code

Ref country code: FR

Ref legal event code: ST

PG25 Lapsed in a contracting state [announced via postgrant information from national office to epo]

Ref country code: DE

Free format text: LAPSE BECAUSE OF NON-PAYMENT OF DUE FEES

Effective date: 20020101